Testing the capital asset pricing model (CAPM) on S&P 500 companies

In this paper our purpose is to test the efficiency of the very popular capital asset pricing model (CAPM). We examine the degree to which CAPM can interpret stock returns. In order to obtain safe results, we choose to focus on a sample of very robust and very important firms. We feel that by selecting as our sample the companies that are included in the S&P 500 index of the US stock exchange, we will have a sample that meets our criteria.

More specifically, for the period 2002 – 2014 and for each individual year, we estimate the stock returns using CAPM. We compare these estimates with actual returns and we compute the deviation. The deviation is a measure of the efficiency of the CAPM model. Less deviation means that the model is more efficient. Based on the annual results, we examine how efficient the model is and also how the efficiency of the model changes year to year over the above period.

Additionally we examine how the efficiency of the model changes between the periods before and after the financial crisis in 2008. We try to understand how this financial crisis event made stock returns more or less unpredictable. Also if there is a change after the financial crisis event, we will examine whether the indications suggest that the change is permanent or is only transient. Based on these results, we will formulate a conclusion about the relationship between stock returns and market returns, as well as the reliability of the CAPM today.